platinotype, cyanotype, Van Dyke, neural networks

The aim of the article is imitation of handmade original techniques used for platinotype, cyanotype, and Van Dyke via digital filters based on artificial neural networks. The proposed methodology of editing information in graphical data, which aims to create a faithful copy of the manual process of alternative photo-graphic techniques, uses backpropagation neural networks, and contributes to the resulting graphics on the basis of defined transformation matrixes. The core of the proposed methodology is the composition of the results generated by individual neural networks with their configurations after adaptation over the proposed training set. An essential part of this article is to verify the proposed methodology in an experimental study.